MCP BigQuery Server

A Model Context Protocol (MCP) server implementation for Google BigQuery, enabling secure, read-only access to BigQuery datasets. This allows Large Language Models (LLMs) to directly query and analyze data through a standardized interface.

• Source: GitHub Repository
• Category: database-messaging-mcp-servers
• Tags: mcp, bigquery, database, cloud

Features

• Implements the Model Context Protocol (MCP) for AI-database communication.
• Provides secure, read-only access to Google BigQuery datasets.
• Enables LLMs (such as Claude) to interact with BigQuery data via a standardized API.
• Allows natural language queries to be translated into SQL and executed on BigQuery.
• Supports integration with Claude Desktop for AI-assisted data analysis.
• Offers installation via Smithery or manual setup.
• Supports authentication with Google Cloud via multiple methods, including service accounts.
• Provides command line arguments for configuration.
• Developer setup instructions for customization or local development.
• Open source under the MIT License.

Limitations

• Currently only supports read-only operations on BigQuery datasets.
• Designed primarily for use with Claude Desktop as of the current developer preview.

Pricing

• Open source and free to use (MIT License).